Live data from Hacker News

The story of Augur, an Ethereum prediction market

decryptmedia.com

1–10 of 62 posts

Re: The story of Augur, an Ethereum prediction market

#8
post #6

Last I read, REP gave up on plans for a decentralised Oracle. That’s the hardest part of this whole deal,unfortunately

That's mostly true, with a caveat. By default the oracle is not decentralized, but what is decentralized is the "backstop". If I create a market, I can also be the one to report on the outcome of the market (versus it being decentralized in the original plan). What they introduced is the concept of a "dispute". So if I report on a market you can dispute it and say "Hey you are lying" and if enough people agree that I lied, you get a portion of the market fees. So you are correct, it's no longer decentralized, but it's also not centralized. Not perfect, but still pretty good.

Re: The story of Augur, an Ethereum prediction market

#9
post #6

Last I read, REP gave up on plans for a decentralised Oracle. That’s the hardest part of this whole deal,unfortunately

That's mostly true, with a caveat. By default the oracle is not decentralized, but what is decentralized is the "backstop". If I create a market, I can also be the one to report on the outcome of the market (versus it being decentralized in the original plan). What they introduced is the concept of a "dispute". So if I report on a market you can dispute it and say "Hey you are lying" and if enough people agree that I…

> if enough people agree that I lied, you get a portion of the market fees

This creates a host of post-closing political incentives. You are no longer betting on an event outcome, but on a consensus process.

Re: The story of Augur, an Ethereum prediction market

#10

Earlier quoted context omitted.

That's mostly true, with a caveat. By default the oracle is not decentralized, but what is decentralized is the "backstop". If I create a market, I can also be the one to report on the outcome of the market (versus it being decentralized in the original plan). What they introduced is the concept of a "dispute". So if I report on a market you can dispute it and say "Hey you are lying" and if enough people agree that I…

> if enough people agree that I lied, you get a portion of the market fees This creates a host of post-closing political incentives. You are no longer betting on an event outcome, but on a consensus process.

It's consensus all the way down: ultimately a cryptosystem is what the majority of the voting/staking/hashing nodes say it is. Remember the DAO reversal?
Post reply on HN